What’s covered in this session:

  • Introduction to Everybody Makes The School.
  • Introduction to Data Science.
  • Hands-on Activity: Identifying Patterns in Data.
  • What is “Everybody Makes The School”.
  • Hands-on Activity: Data Science Challenge.

Everybody Makes The School

Everybody Makes The School is an exciting collaboration between The University of Edinburgh and Newbattle Community High School.

A school is made of not just bricks and mortar, but the people inside it. The students, teaching staff, and support staff at Newbattle Community High School all make a school community we can be proud of.

What do you want to tell people about your school community? Just like a school badge, a school motto, a school song – this is a way you can express what it means to be part of the school.

We used digital technology and data skills to show that everyone matters to the school, and that everyone makes an impact on it. We celebrate our school community through data.Learners gathered and processed data related to the school community – it’s beating heart. They analysed this data and considered playful ways to display it – through digital graphics, smart lighting, or sound. Learners created their own piece of public data-driven digital art to showcase what it means to be part of a Digital Centre of Excellence.

This project doesn’t focus on teaching programming or “digital making”, instead it’s the skills around data science (and the digital skills needed to capture, clean, and process that data) and the creative ways the results can be visualised.

The data produced in a school environment changes on a day-to-day basis, it’s reflective of the people in the school at that time, therefore digital art is dynamic. As the data changes so does the artistic representation.

Learners chose something they want to show or share about the school. They then chose a way to measure or collect the relevant data.  Finally they found a way to visualise it. The group worked in three teams to create data-driven digital art in three ways:
Rather than school motto: experimenting using graphical visualisation
Rather than a school song: experimenting using data with sound
Rather than a school badge: experimenting using data with smart lighting

The success of this project relies in its replicability. If other schools or places or learning want to run their own version they should not have to rely on it being delivered by computing teachers / technology specialists. The costs involved are low, not relying on expensive equipment or high speed Internet access.

 

Session 1: Introduction to Data Science

Resources:

Get the slides with teacher notes.
Get the learners worksheet.

Introduction to Data Science

Start with an ice-breaker. Go around the class and ask learners their name and their favourite app, game, or website. Encourage them to give an answer that wasn’t given before.

Give Spotify as an example. A popular feature of Spotify is Discover Weekly. Each week it makes me a new playlist based on songs it thinks I’ll like.

But how does it know what you like? How does Spotify know if you enjoy a certain song? Do you play it a lot? Have you added it to playlist? Did you share it with friends?

What might I do if you didn’t enjoy a song? Skip it? Delete it?

Explain that Spotify uses a data analytics process to determine the type of music I might like:

Gathering the tracks I like
Exploring the patterns in those tracks
Predicting other tracks I might like

Hands-on Activity: Identifying Patterns in Data
Ask the learners to discuss in pairs (2 mins) the ways Spotify might determine what type of music I like. (Gather)

Once it gathers all the songs I like – it tries to find patterns (things that they all have in common). Encourage learners think about the types of songs they listen to. What sort of things do they have in a common: same style of music, same artist, same album?

Ask the learners to discuss in pairs (2 mins) the ways Spotify might determine what type of music I like. (Explore the patterns)

Once Spotify has found patterns in the data it then predicts other songs I might enjoy and adds them to my Discover Weekly playlist. Using patterns in the listening data, how might it choose to predict (or guess) other songs I might like?

Ask learners to discuss in pairs one way they might use a pattern to suggest another track. (For example, recommend a track from the same artist, recommend a track from the same genre). (Predict)

One of the reasons Spotify works so well at making predictions is they don’t just know what I’m listening to, they know what everyone is listening to. Millions of people use Spotify each month, so they can use all our data together to create better results.

So imagine, you’ve been paired with another user in Brazil. You’ve been paired because you both like a lot of the same music – even though you’ve never met this person Spotify knows you like the same music. Except maybe there is an album they’ve been listening to a lot recently that you’ve never heard. Spotify might infer that, since you like the same sort of music as that person, you’ll probably like this album too and recommends it to you.

But why are Spotify going to all this effort? Get learners to think about the benefits to giving good recommendations to their customers from a business point-of-view  (better service, more customers, willing to pay more money).

Ask learners what other companies do they know that do this too? How does it benefit those companies? What is the experience like for customers?

(Netflix, Amazon, Instagram)

What is “Everybody Makes The School”

So companies that hold data on their customers: it lets them learn about their customers, it let’s them make new products that suit their customers, it might help them solve problems. Just like companies hold data, schools hold data about us.

Learners should consider what data the school might hold on them already (names, ages, attendance, number of school lunches bought) and what data could be measured and gathered? (room temperatures, noise levels in corridors, movement in corridors, happiness of learners)

The slides give examples of the types of data that could be collected, and then interesting ways of presenting it:

For example, the attendance records of each year group could be converted into a symphonic string tune –  a new soundtrack for the school. Notes created by data from S1 and S2 have a higher pitch than those in the upper school. The longer the “streak” of perfect attendance, the longer the note. This composition uses real attendance data. Learners must consider how this data must be stored and cleaned to prevent identification.

Or, a “mood cube” that reflects the current activity level of the school.  The light is brighter the more people are moving around the school. The colour of the light changes depending on what sort of activity is taking place (class time, break time, assembly time). This could be made using Philips Hue bulbs and Python code.

Hands-on Activity: Data Science Challenge
How many smarties are in a tube?
How many of each colour of smarties are in a tube?

Learners can use data analytics (Gather -> Explore -> Predict) to make their own predictions to answer these questions.

Learners should work through the steps on Worksheet 1. To create the visualisations learners could use a spreadsheet.

You’ll need to have lots of little boxes of smarties (or other sweets) – the funsize ones work well. At the end of the lesson use an unopened packet of smarties to see how accurate the predictions have been. How might they improve their predictions? (More data?)

 

Getting Feedback from Learners

When it comes to collecting data about learners it is important to be open and transparent. After all, the workshops explore the power and value of data and how it might be used.

I added a disclaimer at the start of the feedback form. This let them know all questions were optional, and all responses were anonymous. I also made it clear the data collected would be shared with the project partners: University of Edinburgh and Newbattle High School.  Your feedback form should be adjusted to suit your needs and uses.

I chose to use a paper feedback form rather than a digital version. I felt this was quicker to fill in rather than asking students to navigate to a website. The downside was I had to type the results into a spreadsheet manually, but that didn’t take long as it was only a small number of forms.

On the feedback form I only asked 4 questions:

  • Circle the emojis that best describe how you felt about the session
  • What did you learn today that you didn’t know before?
  • What did you enjoy most about the session?
  • What did you not enjoy? What would you change?

The responses gave me enough evidence to gauge if the workshop had achieved its objectives. Did they enjoy it? Did they learn something new? The responses also offered an insight into ways it could be improved.

You can download the feedback template here as a Word document, or as a PDF.

Author: Craig Steele, Digital Skills Education –https://craigsteele.com/

Posted on:

21.01.20